Parking Assistance Boundary Detection for Road Mark Filtering
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing parking assistance devices struggle to accurately detect boundary lines of parking slots due to erroneous detection of road surface marks, leading to inaccurate parking target positions.
Innovation Solution
A parking assistance device that includes a captured image acquisition unit, end detection unit, position information calculation unit, boundary line determination unit, and removal unit to identify and exclude road surface marks within a predetermined range from the detected boundary lines, ensuring accurate boundary line detection and parking guidance.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If boundary line detection is performed using conventional methods, then detection speed is improved, but detection accuracy deteriorates due to erroneous detection of road surface marks
Solution Approach 1:
The patent segments the detection area into multiple regions (first detection area and second detection area) separated by a reference line. The first detection area is for objects closer to the vehicle, and the second detection area is for objects farther away. This segmentation allows different detection strategies for different regions, improving both speed and accuracy by processing areas with different characteristics separately.
Solution Approach 2:
The patent applies different detection criteria and weightings to different regions of the detection area. The first detection area (closer to vehicle) and second detection area (farther from vehicle) are treated with different parameters, allowing the system to optimize detection accuracy for each specific region while maintaining overall efficiency.
2Adaptability or versatility
If all detected ground objects are treated as potential boundary lines, then detection coverage is improved, but false positive rate increases due to road surface marks
Solution Approach 1:
The patent extracts and separates road surface marks from actual boundary lines by using the reference line to divide the detection area. Objects in the first detection area (closer to vehicle) are treated differently from those in the second detection area (farther from vehicle), allowing the system to identify and exclude false positives while maintaining detection coverage.
Solution Approach 2:
The reference line acts as an intermediary element that separates the detection area into two distinct regions. This intermediary structure allows the system to apply different interpretation rules for objects on different sides of the reference line, improving reliability by filtering out road surface marks that would otherwise be mistaken for boundary lines.
Data Source
AI summary
Provided is a parking assistance device including a captured image acquisition unit imaging a road surface around a vehicle and an end detector detecting an end of a ground object on the road surface within a detection area in the captured image. A position calculator calculates position information indicating a positional relationship of the end of the ground object. A boundary line determiner determines whether or not a ground object having the end is a boundary line that defines a parking slot for the vehicle. A removal unit determines whether or not a rear end of a ground object determined as the boundary line by the boundary line determination unit is included in a removal range, determines that a ground object determined as the boundary line is not the boundary line when the rear end is included in the removal range, and excludes the ground object from boundary lines.